skip to main content skip to footer

Families as Facilitators of Children's Intellectual Development at Three Years of Age: A Causal Analysis

Author(s):
Laosa, Luis M.
Publication Year:
1981
Report Number:
RR-81-45
Source:
ETS Research Report
Document Type:
Report
Page Count:
65
Subject/Key Words:
Child Development, Cognitive Development, Family Environment, Hypothesis Testing, Parent Influence, Socioeconomic Influences, Young Children

Abstract

In this paper I deal with what is perhaps the most critical conceptual and methodological issue in research on families as learning environments-the issue of causal inferences. The central concern in this paper is the testing of research hypotheses that involve causal relationships among variables. Practical and ethical considerations in family research make it extremely difficult to assess causation. Advances in causal-inference methodology now make it possible, however, to test theoretical models that involve complex networks of causation. These developments have yielded analytic procedures that broaden the range of possible approaches to the problem of causation. Taking advantage of these technical advances, I elaborated and empirically tested a causal model of families as facilitators of their preschool-age children's intellectual development. To this end, I employed a measurement strategy that included both process and status variables and that obtained assessments by a combination of direct observations in the home, interviews, and psychometric tests. The intent was to construct a reasonably comprehensive causal model incorporating a number of variables that, based on prior research claims, should play an indirect or direct causal role in shaping young children's intellectual development. I accomplished this by means of an approach to path analysis that, unlike other path-analytic techniques with their problems in attenuation from measurement error, allows for the use of hypothetical, error-free unmeasured variables, takes errors of measurement into consideration, and relies on the simultaneous estimation of parameters in the model. (65pp.)

Read More